The rationale of this clinical study is that advanced gene transfer techniques may benefit patients suffering from Canavan Disease (CD), a childhood neurodegenerative disorder with a well-defined genetic basis. CD is a monogeneic, autosomal recessive disease in which the gene coding for the enzyme aspartylacylase (ASPA) is defective. The lack of functional enzyme leads to an increase in the substrate molecule, N-acetyl aspartate (NAA), which prevents normal myelination and results in spongiform degeneration of the brain. The natural history of untreated CD is irreversible brain damage and death within the first decade of life. No effective treatment currently exists for CD, but gene transfer has the potential to arrest or reverse the course of this otherwise fatal disease. We propose to surgically administer the aspartylacylase (ASPA) gene directly to affected regions of the brain, using recombinant adeno-associated viral (AAV) vectors. In a previous Phase I study for CD using a first-generation vector known as LPD - comprised of an AAV-based plasmid in conjunction with a condensing agent (protamine sulfate) and a liposomal formulation (DC-CHOL/DOPE) - surgical gene transfer to the CSF space of the brain was shown to be safe. In addition, gene transfer was associated with regionally lowered NAA and various functional neurological improvements. These positive changes were subtle in some patients and the more impressive changes that occurred were relatively transient, which we believe was primarily due to inadequacies of the vector and delivery system. Gene transfer technology has advanced considerably in the last five years and based on more recent work, we have determined that AAV vectors are more efficient and safer than LPD as a gene delivery system, especially when used in conjunction with an intraparenchymal approach. Pre-clinical experiments have been performed which suggest the superiority of AAV as a tool for clinical gene transfer. Current helper-free AAV packaging and purification methods eliminate the possibility of wild-type virus recombination and provide highly purified vector titers exceeding 1012 genomic particles/ml. The AAV vectors that we have tested in rodents and primates contain a mammalian neural promoter element with additional postregulatory elements, which are able to drive high levels of long-term transgene expression. The AAV vector containing human aspartylacylase cDNA (AAV-ASPA) will be neurosurgically injected into defined subcortical regions of the brain in a cohort of patients with Canavan Disease. The pre- and post-delivery assessments will involve non-invasive biochemical, radiological, electrophysiologic, neurological, and cognitive tests. These outcome measures - NAA levels, MRI signal changes, evoked potentials, and standardized neurodevelopmental assessments - will be systematically recorded and statistical analyses will be performed. We anticipate that the meticulous collection and analysis of data from this study will provide the basis for a larger (Phase II/III) clinical trial, and that this study will be useful as a reference for all future trials using viral vectors in the human brain.